A New Fuzzy-Cluster-Based Cycle-Slip Detection Method for GPS Single-Frequency Observation

• 发布时间：[2022-02-18] 来源：[学院] 点击量：[30708]

作者: Li, ZN (Li, Zongnan); Li, M (Li, Min); Shi, C (Shi, Chuang); Chen, L (Chen, Liang); Deng, CL (Deng, Chenlong); Song, WW (Song, Weiwei); Liu, RL (Liu, Renli); Zhang, PH (Zhang, Peihua)

来源出版物: REMOTE SENSING  卷: 11  期: 24  文献号: 2896  DOI: 10.3390/rs11242896  出版年: DEC 2 2019 
摘要: The development of low-cost, small, modular receivers and their application in diverse scenarios with complex data quality has increased the requirements of single-frequency carrier-phase data preprocessing in real time. Different methods have been developed, but successful detection is not always ensured. The issue is crucial for high-precision positioning with Global Positioning System (GPS). Aiming at a high detection rate and low false-alarm rate, we propose a new cycle-slip detection method based on fuzzy-cluster. It consists of two steps. The first is identification of the epoch when cycle slips appear using Chi-square test based on time-differenced observations. The second is identification of the satellite which suffers from cycle slips using the fuzzy-cluster algorithm. To verify the performance of the proposed method, we compared it to a current robust method using real single-frequency data with simulated cycle slips. Results indicate that the proposed method outperforms the robust estimation method, with a higher correct-detection rate and lower undetection rate. As the number of satellites simulated with cycle slips increases, the correct-detection rate rapidly decreases from 100% to below 50% with the robust estimation method. While the correct-detection rate using the proposed method is always more than 60%, even if the number of satellites simulated with cycle slips reaches five. In addition, the proposed method always has a lower undetection rate than the robust estimation method. Simulation showed that when the number of satellites with cycle slips exceeds three, the undetection rate increases to more than 30%, reaching 70% for the robust estimation method and less than 30% for the proposed method.